Just Call Me Egghead

The surprising smarts of crows and jays are forcing scientists to reconsider how they define intelligence.

By Charlotte Hu|Monday, August 01, 2011
shutterstock_105559973
shutterstock_105559973
The western scrub-jay plans for the future, remembers the past, and speculates about what its fellow jays are thinking.
Anatoliy Lukich / Shutterstock

Nicky Clayton is no better at sitting still than are the birds she studies. Back in the 1990s, her colleagues at the University of California, Davis, would stay at their computers at lunchtime, but she would wander outside and watch as western scrub-jays stole bits of students’ meals and secretively cached the food. During these informal field studies, Clayton, an experimental psychologist, noticed that the birds returned frequently to their stashes and changed their hiding places.

“I thought, ‘This is odd,’ ” she says. “I assumed birds would cache for a long time—days or months. But this was for minutes.” She theorized that the birds were moving their caches to avoid pilfering. When food was plentiful, they grabbed as much as possible and hid it, then hid it again when they could do so without being observed by potential thieves. That behavior implied that the scrub-jays might be thinking about other birds’ potential actions, a type of flexible thinking that was supposedly beyond the capabilities of a scrub-jay’s little brain.

Clayton realized that if she could capture this caching behavior in the laboratory, she might be able to decode the social cognition of birds—the way they think about one another. She might learn whether they are capable of deception, if they respond differently to individual competitors, how well they evaluate their degree of privacy, and other aspects of their mental processes. “I had a lucky break with caching,” Clayton says now. “I saw this as a niche, an area that other people weren’t busy with that might be quite interesting. Little did I know where it would lead.”

Scientists had already established the amazing memories of corvids, the family of birds that includes jays, crows, ravens, and nutcrackers. The Clark’s nutcracker can hide thousands of seeds at a time and has passed tests of recall up to 285 days later. Clayton sought to find out how deep those skills run. Many animals have impressive mental capabilities for certain narrow tasks, but such aptitudes seem to reflect hardwired or conditioned adaptations to specific challenges. That is distinctly different from a human’s ability to create and manipulate a flexible mental model of the world.

Within a few years of her lunchtime insight, Clayton was conducting the first experimental demonstrations of a non­human animal engaged in mental time travel. Her experiments demonstrated that scrub-jays plan for the future, recall incidents from the past, and mentally model the thinking of their peers. Since then her work has expanded even further. She has found other mental capacities in birds that rival or surpass those of any other nonhuman species and come uncannily close to abilities we thought were ours alone.

When she left her native england in 1995 for her position at U.C. Davis, Clayton already believed there was much more to corvid cognition than people thought, having studied their memory development during her graduate studies at Oxford University. The scrub-jays in the park were not the only thing that lit a fire under her. She also drew inspiration from another English expatriate, a neuroscientist doing related work from a different perspective, studying social cognition in rhesus macaques. That researcher, Nathan Emery, would later become her husband. Like many romances, theirs faced some early tests. Soon after they met, Clayton recalls critiquing a draft paper of Emery’s on primate eye-gazing. The paper included a chart listing mental capabilities that were the sole province of primates. “I kept writing in the margin, ‘Oh, no it’s not,’ ” she says, “which I’m sure he found very irritating.”

After Clayton’s groundbreaking work at U.C. Davis, she returned to England in 2000 and rapidly rose to become a full professor at Cambridge University and director of natural sciences at the university’s Clare College. Emery made the move as well, becoming a Royal Society University Research Fellow at Cambridge. (He is now senior lecturer in cognitive biology at Queen Mary University of London.) They married in 2001 and together have pursued the study of animal social cognition, with Clayton drawing her husband’s attention more to the avian side of things.

As she waded into the study of complex cognition, Clayton found herself in a field that was full of intriguing but poorly documented research. For decades scientists sought to demonstrate that nonhuman primates have mental abilities similar to ours. Experiments teaching chimps like the famous Washoe to communicate using sign language received enormous publicity but rarely survived critical analysis. Other projects aimed to show that animals have a theory of mind, the ability to model the thinking of others (as when we judge whether a poker player is bluffing or whether a potential mate is truly in love).

But here, too, the studies seemed to glimpse such abilities in animals without ever delivering definitive proof. Experimental psychologist Sara Shettleworth of the University of Toronto summed up the shaky history of these studies in a textbook in which she reported a history of ambiguous and mistaken findings in the field, concluding that most of the claims about animal social cognition in the last century were inadequately proven and needed further testing.

The earlier research “was not ecologically inspired,” Clayton says. “It was psychologically inspired. We were asking, ‘How can we understand a chimpanzee mind?’ while the mind we really understand is the human’s.” From the start, she saw the problem differently. “If theory of mind means thinking about how others are thinking, then how you think as a human might differ from how you think as a scrub-jay or an ape,” she says. Instead of trying to train animals to do human tasks, she studied mental adaptations that corvids might need in their own setting. And instead of relying on field observations or working with a single trained subject, she conducted repeatable laboratory experiments carefully designed to rule out alternative interpretations.

At first Clayton’s captive scrub-jays refused to engage in the caching behavior she had seen on the Davis campus. She released the birds into a room with food and plenty of places to hide it, but when she returned the birds to their cages, nothing had been left behind. In the wild, she realized, scrub-jays cache mostly in their home territory. When Clayton allowed the birds to cache in the enclosures where they lived day to day, they quickly began storing the worms and nuts she provided just as they had the booty stolen from students’ lunches.

In collaboration with Tony Dickinson, a comparative psychologist at Cambridge, Clayton showed in 1998 that the remarkable cognitive capacities of scrub-jays include the ability to track the passage of time. She found that the birds would return to caches when the food they had hidden was about to spoil. The jays also adjusted their retrieval pattern when presented with new information about how quickly a certain food goes bad, abandoning those caches with contents past their expiration date.

Jays even prepare for the future. Given the opportunity in the evening to place a cache in either of two cages—one in which they had previously been hungry at breakfast time and one in which they had previously been fed—the birds made the correct choice, provisioning the cage where breakfast had not been provided in the past. Neuroscientists believe that episodic memory—the recall of a moment rather than a skill—relies on the same structures in the human brain’s hippocampus as does imagination. Both functions demonstrate our capacity for mental time travel, the ability to recall past events or envision new ones. Clayton’s experiments raise for the first time the possibility that scrub-jays can mentally time travel too. “We thought these abilities were uniquely human,” she says. “The fact that jays have them says no.”

In Clayton’s experiments, the scrub-jays’ social thinking proved more complex than anyone had predicted. The birds remembered if they were being watched by other birds when they cached, and by which ones. They would wait until a potential thief was distracted before hiding food, or would choose a spot that was dark or otherwise difficult for the competing bird to see. If another bird could potentially hear the process of hiding the food, they chose quieter material in which to dig—sand rather than pebbles. If they had no choice but to cache in plain sight, the scrub-jays would return soon after, when conditions permitted privacy.

Clayton recognized that the birds’ behaviors might be innate. To address this possibility, she, Emery, and colleagues did other experiments. They hand-raised scrub-jays without giving them the opportunity to steal from other birds’ caches. Those “naive” jays did not take precautions to avoid being victims of theft. Apparently, the ability to avoid theft by others depended on projecting a bird’s own experience. It took a thief to know a thief.

“The fact that it was only experienced thieves that did it—that really blew my mind,” Clayton says. It was at that point that she and Emery began to invoke cognition to explain the birds’ maneuverings.

Other researchers have also narrowed the mental differences between people and birds. Bernd Heinrich, a behavioral ecologist at the University of Vermont, has documented ravens’ extraordinary social organization, secretive storing of food, elaborate communication, and extensive play, both in the field and in an aviary. He has described these results in books such as Mind of the Raven.

Recently, in collaboration with Thomas Bugnyar, an Austrian biologist, Heinrich measured ravens’ caching and problem solving in the lab, with results similar to Clayton’s. The two researchers conclude, even more unequivocally than she, that ravens attribute to their competitors “the capacity of knowing.” Other researchers studied the cleverness of New Caledonian crows, which can spontaneously invent novel tools.

Looking beyond corvids, some animal behaviorists have examined how songbirds use grammar. And some have ventured further down the evolutionary tree. One study attributed higher mental functions to fish, presenting evidence that African cichlids can reason inferentially. The accelerating stream of discoveries is challenging the conventional view of what animal minds can do.

Clayton’s work also has its critics—not surprising, given the century’s worth of unproven claims regarding animal intelligence. Shettleworth suggests that Clayton and Emery need to repeat the thieving experiment in a different way, with a large number of fresh birds divided into groups according to whether or not they have experience as thieves. “I think the work is provocative but not proven,” Shettleworth says, “because those birds have a history.”

Daniel Povinelli, a University of Louisiana biologist, contends that in regard to higher-order thinking, there is still more evidence for human-animal differences than for similarities. Our own theory of mind fools us into seeing our abilities in animals, even when simpler explanations would suffice, he says. Povinelli does not agree that Clayton’s experiments show birds can mentally place themselves in different times as humans do. “They’re just representing, as Nicky has elegantly shown, that they can keep track of the relative breakdown of foods in different locations,” Povinelli says. Mechanical clocks keep excellent time without having a notion of what time is, he points out. Nobody disputes that animals have exquisite internal clocks, but that does not necessarily mean they have a concept of time. “Do they understand the connection between the decaying of insects, the sun setting and rising, the seasons passing, other birds dying? There is absolutely no evidence of this. None. Zero.”

Povinelli also argues that Clayton’s experimental design was inadequate to prove that the birds formulate a theory of mind. A theory of mind requires the ability to mentally model counterfactual ideas, such as thinking about what another person would do in a hypothetical situation. But the birds might be responding to their competitors by using rules, not by imagining the content of their thoughts. As for the hand-reared, nonthieving scrub-jays, they may have missed developmental stages or input that help formulate the rules. “Nicky and those guys want to say the birds are thinking, ‘I know what it was like when the birds stole from me,’ ” Povinelli says. But he believes the naive birds’ failure to anticipate thievery does not support the interpretation.

Emery counters that there is a limit to what we can expect to learn from animal minds. “We will never be able to find human theory of mind in nonhumans. They have their own social cognition that has evolved for their own problems,” he says. “If they had human theory of mind, they would be little humans.” Or, as Clayton says, quoting the philosopher Ludwig Wittgenstein, “If a lion could speak we wouldn’t be able to understand him.”

The predominant attitude of Western science, Clayton says, has been that animals are unthinking automatons unless proved otherwise, in line with the biblical view that God gave mankind the animal kingdom for our use. But she cites a Hindu colleague who took the opposite point of view, putting the burden of proof on scientists to show that animals are not mentally complex. “Why should you start out with the idea that animals don’t have a theory of mind?” Clayton asks. “Why not start out with the idea that they do?”

The risk, she recognizes, is interpreting animal behavior using the mental machinery with which we negotiate our human relationships. A dog owner comes home to find a mess but shows mercy because the dog seems remorseful. But does the dog really feel regret as we would? Is the show of regret a conditioned response associated with receiving a less severe punishment? Or could the animal be manifesting an instinctive program, treating the owner as a dominant member of the pack? Perhaps all three processes are at work.

Clayton has built her career trying to avoid these uncertainties. She believes that only experiments in the laboratory can escape the thicket of alternative interpretations that confound field observations, so her team continues to endure the painstaking and time-consuming task of acquiring and working with lab-bound corvids. Nevertheless, wildlife biologists report tales of bird intelligence outside the lab as well.

In one compelling example, wildlife researcher Stacia Backensto, then a graduate student at the University of Alaska at Fairbanks, was stymied by bird cognition when she began studying how ravens have adapted to human activity on the oil fields of the Arctic coast, even making use of heat escaping from buildings to cope with the darkest, coldest days. “It’s interesting to be studying something so smart. You’re constantly dueling with this bird,” she says. “You’re constantly playing these games to outsmart it.”

Backensto discovered she could get closer to the ravens if she wore an oil field worker’s uniform. Still, she found it almost impossible to catch the birds in the second year of her study. They had learned all her tricks—even ravens she had not seen before, in areas she had not previously visited. Finally she had to don a complete disguise: a uniform stuffed with pillows plus a shaggy wig, fake beard, glasses, and a mustache. It worked, although the university’s business office wanted to know why Backensto was spending research funds at a place called the Party Palace.

Beyond the anecdotes and individual case studies, there is a common thread among the birds that show the strongest signs of intelligence: the ravens, jays, and other corvids, along with parrots. Each of these species possesses an avian neocortex of exceptional size relative to its body, rears its young for an extended period, and lives in a complex social environment—not merely in a large population of cooperating creatures, such as bees or ants, but in a dynamic setting of alliances and competition. The same is true of the most clearly intelligent mammals: toothed whales, dolphins, and primates.

“It’s not just living in big groups, it’s the complexity of social life,” Clayton says. The social hypothesis explains why animals might benefit by becoming smarter. Clayton suggests that a cognitive arms race among their own kind pushed corvids to develop better brains, as the spy-versus-spy game of caching, stealing, hiding, and deceiving escalated the need for an ever sharper mind. Once the ability to think flexibly emerges, descendants can apply it to face varying challenges. For example, ravens and killer whales, both highly social, also both alter the ways they gather food and use their habitat so they can live near the equator as well as in the Arctic.

In 2009 Emery investigated the latent intelligence of corvids by presenting complex tasks requiring tools to rooks, which do not use tools in the wild. With each step in his laboratory experiments, the challenges got harder and more complicated, but the rooks produced solutions without resorting to trial and error, suggesting they understand cause and effect. They chose rocks and sticks to drop down a tube in order to open a door to get food. In an experiment inspired by Aesop’s fables, Emery presented the rooks with a worm floating below reach in a tube of water. The birds put rocks in the tube to raise the water level to capture the worm. They even manufactured tools, bending a wire to make a hook to pull a bucket holding food out of a tube. The tool worked only with a bend of a precise curvature, around 100 degrees. “We wouldn’t have expected that at all,” Emery says. “That’s why we said in the paper that it is an example of insight. It’s coming up with a novel solution, to innovate.”

Birds and mammals are far apart on the tree of life. Their last common ancestor lived 280 million years ago and their brains are quite different in size and structure; birds notably lack the mammalian six-layer cortex. So Clayton and Emery argue that intelligence had to evolve separately in corvids and primates, converging to solve the same problems of managing social interaction. Intelligence might turn up anywhere it aids survival. It may be rare only because it is not needed very often. “We think intelligence is this great thing because it’s the thing that has made us special,” Clayton says. “Yet when you compare us with insects, such as species of mosquitoes, there are a number of measures where we are not the best.”

We cannot say for certain how important thinking is for action in either animals or human beings. Shettleworth concedes that Clayton’s scrub-jays met the behavioral criteria for future planning when they cached their breakfast in the right cage before bedtime, but queries: “Does that mean they are thinking about breakfast when doing it? We don’t know.” Shettleworth notes that even in people, the unconscious connections of associative learning are behind complex behaviors, such as driving. “Conscious cognition may be very much overrated in our conduct of daily life,” she says.

In the end, we cannot be sure of another person’s conscious thinking, much less the thinking of another species. A computer can be programmed to seem conscious. Scrub-jays might be built to seem conscious too. Even a person could falsely claim to be conscious, and we would have a difficult time knowing the difference. Nonetheless, these existential conundrums do not daunt Clayton, who still watches her scrub-jay research subjects with curiosity and optimism. “I just like watching them behave,” she says, “and using that to generate ideas.” 

ADVERTISEMENT
Comment on this article
ADVERTISEMENT
ADVERTISEMENT
ADVERTISEMENT
DSC-CV1218web
+